PolicyEngine Macro
US · official data, no model outlook
The US data, with every number tied to its vintage.
Forecasts is what the platform produces for the US; this is what it is produced from. Five topic pages, one per question, each carrying the same four layers in order: what the data says, what the models see and cannot see, the command that reproduces it, and the dated file every number came from. Compare the UK scope →
No model in this suite forecasts the US economy. The only member whose question types include forecast is boe-svar, and its geography is UK-only. Both US members are shock models and the registry says so in their own words: frb-us records predictive validation not_assessed — “LONGBASE is an illustrative tracking baseline, not an official Fed forecast, and no historical forecast evaluation is published here.” — and us-hank records not_applicable. Where a model view exists at all it is on the topic page that reads the series, in the same breath as the limits that qualify it.
Five topics, each with its data, its models, and its limits.
The rest of this site is organised by model — a reader has to know what a heterogeneous-agent New Keynesian model is before they can find the US inflation response. These pages are the other axis: one page per question, each carrying what the data says, what the models see and cannot see, the command that reproduces it, and the dated file every number came from. This page is the way in, and the provenance behind them.
Five topics here, six on the UK side. The missing one is public finances, and it is missing for a reason worth stating: the vintage store holds no US fiscal series — no federal balance, no debt-to-GDP — so that page would open with no number, no observation period and no vintage. Nor is there a model to put beside one: obr-macro is uk-only, and neither US model reports a fiscal balance among its outputs. Filling it needs two things in this order — the FRED series added to the append-only fetcher so vintages accumulate, and a US model that answers a fiscal question. Until then the tab is absent rather than empty.
Growth, inflation, and unemployment over time.
The topic pages carry the current reading and, where one exists, the conditioning baseline beside it; these three charts are the one place the whole path is drawn. Each is generated from the same committed point-in-time observations, and the source link opens the corresponding official series. Two of the three are shorter than their UK counterparts: GDPC1 is quarterly, so twenty points span 2021Q3–2026Q2, but CPIAUCSL and UNRATE are monthly, so the same twenty-point window covers only 2024-11–2026-07 and 2024-11–2026-07. And no US series in this store starts before 2020-01, where the ONS series behind the UK charts reach back to 1955Q1: the long view here is as long as the store, not as long as the published record.
Every official series, and the topic that reads each one.
This is an index, not a second dashboard. The readings, their period comparisons and the model view of each one are on the topic pages, beside the limits that qualify them; what is here is what a reader needs in order to check a number — what is tracked, which topic carries it, the period it covers, and the dated snapshot this site read.
One column the UK hub carries is missing here, and the absence is the point: the ONS stamps every series with the date it was released, and FRED stamps none of them through this site's fetcher. Rather than a column of cells reading “not supplied”, it is gone, and the stored vintage — the day this site fetched the snapshot and committed it — is the only date on the row an artifact stands behind.
| Series | Topic | Observation | Stored vintage | Source |
|---|---|---|---|---|
| Real gross domestic product | Growth | 2026Q2 | 2026-08-04 | FRED · GDPC1 |
| Consumer price index, all urban consumers | Inflation | 2026-07 | 2026-08-26 | FRED · CPIAUCSL |
| Unemployment rate | Jobs | 2026-07 | 2026-08-10 | FRED · UNRATE |
| Nonfarm payroll employment | Jobs | 2026-07 | 2026-08-10 | FRED · PAYEMS |
Rates & markets — the policy rate and the 10-year Treasury.
A monthly average of the effective federal funds rate and a business-day constant-maturity quote, neither with an announced release date. Where the UK hub indexes four points on the gilt curve, this store holds one Treasury maturity, so there is no US curve to put beside it. No model on this site forecasts the funds rate or a Treasury yield: the rates and Treasuries topic carries the levels and quotes, from the capability registry, why the honest answer there is a refusal rather than a number.
| Series | Topic | Observation | Stored vintage | Source |
|---|---|---|---|---|
| Effective federal funds rate | Rates and Treasuries | 2026-07 | 2026-08-04 | FRED · FEDFUNDS |
| 10-year Treasury yield | Rates and Treasuries | 2026-08-24 | 2026-08-26 | FRED · DGS10 |
No calendar to publish, and where the old numbers are kept.
The UK hub ends on a release calendar, grouped by release date, because every ONS series arrives with an announced next release in its metadata. There is no US counterpart to it. FRED returns neither a release date nor a next-release date through this site's fetcher, and every series above stores both fields as null; BEA and BLS publish their schedules elsewhere, but copying those dates here would put a date in front of a reader that no committed artifact stands behind. The band that would carry the calendar says this instead.
BEA, BLS and Federal Reserve series are distributed through FRED. Official statistics are revised: every number in this section — this page and the five topic pages alike — is read from committed snapshots in the append-only vintage store; old snapshots remain available, so a number published here can be reproduced against the data as it stood when it was published.
Every series above is published as dated, immutable JSON snapshots — browse the data store, its release calendar, and the as-of recipe →
These figures are the backdrop the US models run against — the outturns a scored reform starts from, and the fixed path an frb-us shock is measured against. None of them is scored on the forecast record: scoring needs a forecast, and no model here makes a US one.